4.5 Methodology 94
4.5.3. Model Identification 97
The nature of SVAR requires imposition of enough restrictions so as to identify the orthogonal structural components of the error terms that are present in the shocks. Note that this is at variance to the standard recursive Cholesky orthogonalisation. The non- recursive orthogonalisation of the error terms produced through this process is used for the impulse response functions and variance decomposition.
For clarity, assume that π¦π‘is comprised of vector of endogenous variables. For example, say kth element of endogenous variables in our model where β πΈ[π£π‘π£Μπ‘] is the residual of the covariance matrix. Therefore, our identification procedure follows:
π΄π£π‘ = π΅ππ‘ (4.7)
Where π£π‘and ππ‘ are vectors with lag length k, π£π‘is the observed residual and ππ‘ represents the unobservable structural innovations. A and B are k x k matrices which are to be estimated. However, innovation ππ‘ is assumed to be orthogonal in nature. Hence the covariance is an identity matrixπΈ[ππ‘ππ‘π‘]=I. Imposition of restrictions on A and B is made possible due to the orthogonal assumption of ππ‘. hence we have:
π΄ β π΄Μ=π΅π΅Μ (4.8) The link between the reduced form and the structural form of the VAR model is presented as follows:
π΅(πΏ) = π΅0+ π΅+(πΏ) (4.9) π΄(πΏ) = βπ΅0β1π΅+(πΏ) (4.10)
β =Μ π΅0β1π΄π΅0β1 (4.11) Equation 6.9 is the structural form divided into contemporaneous correlations i.eπ΅0 and π΅+(πΏ).The former represents correlations at lag zero while the later represents correlations at all strictly positive lags. Equation 4.10 separates each reduced form coefficient into its structural counterpart π΅0, identified through the reduced form,β =Μ πΈ[ππ‘ππ‘π‘], and the diagonal covariance matrix of the structural form, π΄ = πΈ[π£π‘π£Μπ‘] as shown in 4.11.
Furthermore, due to the vulnerability of long run restrictions to serious misspecification problems, we use a contemporaneous restriction on theπ΅0 matrix to identify the shocks as shown in equation 4.12 since this study is interested in short-run and medium term responses (see Leeper, Sims and Zha, 1996; Elbourne, 2007).
[ π£π‘ππππ
π£π‘πππππ π£π‘πππ‘π π£π‘ππ ππ
π£π‘ππππ π£π‘ππ₯π π£π‘πππ π£π‘πππππ]
=
[
1 0 0 0 0 0 0 0
π΅210 1 0 0 0 0 0 0
π΅310 0 1 0 0 0 0 π΅380 π΅410 0 π΅430 1 0 0 0 0
0 0 π΅530 π΅540 1 π΅560 0 0 π΅610 π΅620 0 0 π΅650 1 π΅670 0 π΅710 π΅720 π΅730 π΅740 π΅750 π΅760 1 0
0 0 0 0 π΅850 π΅860 0 1 ] [ ππ‘ππππ
ππ‘πππππ ππ‘πππ‘π ππ‘ππ ππ
ππ‘ππππ ππ‘ππ₯π ππ‘πππ ππ‘πππππ]
(4.12)
There are eight variables in the SVAR model namely oil price (poil) which is the exogenous variable.It occupies row 1 and it puts external pressure on the economy.
Endogenous variables are arranged as follows: oil resources growth rate (oilgr); interest rates (intr); money supply growth rate (msgr); inflation rate (inf); exchange rate (exr);
manufacturing output growth (mgr); and GDP growth rate (gdpgr). The role assigned to each variable is explained in the flow chart in figure 4.6
The oil price is viewed as the external shock to the entire system, meaning that it affects the monetary policy transmission mechanism MTM. The oil output growth rate is included in the MTM based on the controversy that often in oil exporting countries economic policies receive shock from the global oil price through the individual countryβs oil output levels. Berument, etal (2004) in his study of Oman and UAE found that oil price shocks affect economic policies of these countries through their output levels. He argued that since these countries are heavily dependent on oil, the influence of oil prices on output is translated to economic wealth which dictates the behaviour of economic policies.
On the other hand, Jemenez and Rodriguez (2005) noted that most oil exporting countries run very open and liberal economies which make their economic policies highly susceptible to external shocks. They argued that as these countries are heavily dependent on oil, fluctuations in the global price of oil affect their economic policies without necessarily passing through their output levels.
The MTM comprises of the monetary policy instruments (MPIs) and the policy variables. The MPIs are interest rates and money supply growth rates. They are the operating targets (see Mahmud, 2009; Elborne, 2007). While the intermediate targets are the policy variables, namely inflation rate and exchange rate. They are viewed as an internal policy shock to the system (see Ushie, Adeniyi and Akongwale, 2012; Mordi and Adebiyi, 2010; Ngalawa and Viegi, 2011). The policy goals used are manufacturing output growth and GDP growth rate. Note that the manufacturing output growth is calculated as a percentage of GDP. The ordering of the variables follows Pesaran and Shin (1998) in order to overcome arbitrary ordering and likelihood of contemporaneous correlation.
The flow chart for the system is explained in figure 6.1. The MTM is the monetary policy transmission mechanism which involves the operating targets; the MPIs which are the interest rates and the money supply growth rates. It also includes the policy variables which are the intermediate targets, that is inflation rates and exchange rates.
Policy goals include manufacturing output growth and GDP growth rate.
The flow chart shows a schematic diagram of the monetary policy transmission mechanism (MTM). Firstly, there are the direct effect of the oil price shock on all the components in the model, which are oil growth rate, MTM, and outputs. Secondly, the diagram shows how the oil price shock passes through each of the components as they are arranged. Thirdly, we focus on the direct influence of the shocks in the MTM on the monetary policy goals; and finally we analyse how shocks in the MTM affect the variables in the model.
Figure 4.6 Flow chart for variablesβ roles in the monetary transmission system
Oil Price Shocks
Oil Output Growth Rate
Outputs
MPIs
Policy Variables
MTM